Data may be the “oil” of the information economy, but without microelectronics, those data are just assortments of zeros and ones. Semiconductors, silicon-based “translators” of data into information, are essential to the 21st century economy, making connectivity and all the resulting “information-related” innovations possible. The United States has dominated this field since the semiconductor was invented over a half century ago and assesses any attempt to replace it as No. 1 as an affront to national pride and a threat to national security.
That mentality first surfaced in the 1980s, when the U.S. mobilized to fend off a Japanese challenge to its supremacy in the industry. It has re-emerged in recent years as China recognized the role that semiconductors play as it aims to become the world’s most innovative nation and to dominate production of high technologies.
The U.S. is playing defense — late last month, Washington imposed new restrictions on the export to China of semiconductor production equipment and other technology — but that isn’t enough. The U.S. needs a more expansive strategy to maintain its lead in this field, and that means working closely with its allies, especially Japan.
The U.S. leads the world in leading-edge semiconductor research, design and manufacturing. The Semiconductor Industry Association reckons that U.S. companies accounted for about half of the $469 billion global semiconductor market in 2018, but rightly adds that lead is not assured. A steady flow of resources is essential to stay on the frontiers of innovation in this industry and those it empowers.
China gets it. It is the world’s largest consumer of semiconductors, gobbling up about 60 percent of the global supply. While the country produces about 90 percent of global production of smartphones and almost two-thirds of personal computers, China imports about 84 percent of the semiconductors it needs, and only half of domestic production (8 percent) is by Chinese firms. Chinese companies produce less than 5 percent of the world’s demand for semiconductors used in consumer electronics.
China’s reliance on imports won’t change anytime soon, but not for lack of trying. In 2014, Beijing adopted the “National Integrated Circuit Industry Development Guidelines,” which set the goal of becoming a global leader in all segments of the semiconductor industry by 2030.
“Made in China 2025,” the national blueprint for mastering and dominating a range of leading-edge technologies adopted a year later, identified advanced semiconductor manufacturing as a core component of that effort. That ambitious project aims to produce 40 percent of China’s semiconductor requirements by 2020 and 70 percent by 2025.
Chinese President Xi Jinping underscored the urgency of that task in 2016 when he said “the fact that core technology is controlled by others is our greatest hidden danger.” Vice Premier Ma Kai was more explicit at the 2018 National People’s Congress when he flatly stated that “we cannot be reliant on foreign chips.”
Beijing has put its money where its ambitions are, investing more than $170 billion through public and private sources in its semiconductor industry; a single government fund has $29 billion. By the end of 2017, just three years after revealing its guidelines, China announced plans to build 14 new chip factories. There has been some payoff: China is now designing chips for artificial intelligence and low-end memory storage.
Despite the push, Chinese industry continues to lag. James Lewis, an expert at the Center for Strategic and International Studies, argues that China’s “relative technological backwardness” and increasing state control over the economy and information will slow innovation. China’s best companies make chips at 14 nanometers while the leading edge is 7 nanometers, prompting analysts to conclude that Chinese manufacturers are two generations behind the industry’s top performers. The U.S. aims to keep that lead.
New restrictions on exports require U.S. companies to acquire licenses to sell certain items to companies in China (and other countries) that support the military, even if the products are for civilian use. The regulations take aim at China’s military-civil fusion, which the U.S. State Department — citing Xi — calls “an attempt to deliberately erase the line between China’s military and civilian sectors.”
U.S. Commerce Secretary Wilbur Ross explained the logic behind the rules, cautioning against “doing business with countries that have histories of diverting goods purchased from U.S. companies for military applications,” but this goes much further than that. Now, companies must consider whether their sales are to a civilian company whose actions are intended to support the operation of a military item. Given the intimacy of public-private relations in China and its government’s ability to direct business behavior, some observers speculate that the new rule is practically a ban on exports in this field.
The U.S. government also proposed a third rule that would force foreign companies shipping certain U.S. goods to China to seek approval from their own governments and from Washington. Since semiconductors are created from complex supply chains, this restriction, which applies to businesses in the U.S. and its allies, would impact Chinese industry and reinforce Chinese determination to build indigenous capacity. But Chinese links are critical pieces of those supply chains, which makes it difficult to just cut them out.
Ultimately, semiconductor manufacturing equipment (SME), not chips, is key. According to a new study from the Georgetown University Center for Security and Emerging Technology (CSET), SME is the “primary and most complex input” in the construction of fabrication facilities, accounting for about 80 percent of construction costs.
Today, three countries claim 90 percent of SME global market share: the U.S., Japan and the Netherlands. The CSET analysis concludes that “if these states deny access to this specialized equipment, China would find it nearly impossible to develop or maintain advanced chip factories for the foreseeable future.” The U.S. administration of President Donald Trump understands the problem: Late last year, its two-year pressure campaign persuaded the Netherlands government to block a $150 million sale of the most advanced Dutch chip manufacturing technology to a Chinese company.
Here again, the U.S. administration must fight its unilateralist instincts if its policy is to succeed. The CSET study concluded that concerted action by the three governments on this supply chain chokepoint “could decisively maintain China’s continued dependence on democratic states for chips at or near state-of-the-art.” The reference to “democratic states” is critical. CSET calls for those three governments, along with Taiwan and South Korea, two leading semiconductor manufacturers, to coordinate export policies on SME and chips.
The CSET conclusions echo those of a study earlier this year by the Center for a New American Security (CNAS), which in addition to endorsing multilateral export controls, called for the creation of an international fabrication consortium, which would include Japan and the Netherlands, to create scale, defray costs and establish a secure supply chain for semiconductors.
Eric Sayers, one of the authors, argues that “if there is an issue of alliance significance, the semiconductor industry is it. Whoever controls state of the art microelectronics has a hold on both future commercial technologies and military applications.”
Sayers added that “what we should learn from the 5G telecommunications competition is that simply admiring an industry we are competitive in and letting China catch up is not a recipe for success. We need to take steps now to remain the leader in microelectronic production so that we aren’t looking back in five to 10 years wishing we had made different decisions.”
The CNAS and CSET reports make clear that blueprints and recommendations are available. And they underscore the need to take those steps with allies and partners if those countries are to retain their position on the leading edge of high technologies.